QTN mapping, gene prediction, and simulation breeding of four-seed pod numbers in soybean

The number of four-seed pods (NFSP) in soybean is an important yield trait and a quantitative trait regulated by multiple genes. Mapping quantitative trait nucleotide (QTN), mining major genes, and screening excellent breeding schemes for NFSP are of great significance for breeding high-yielding soybean varieties. In this study, a germplasm population (GP) containing 455 soybean varieties was planted in five environments to investigate NFSP. Single-nucleotide polymorphism (SNP) genotype data were obtained based on the Axiom_SoyaSNP 180K chip, and a genome-wide association study (GWAS) was used to locate QTNs, which were used to predict candidate genes and simulate breeding. The results showed that there was genetic variation in NFSP and different gene expression in various environments. The broad-sense heritability of NFSP over multiple environments was 72.7%. A total of 89 QTNs correlated with the number of four pods were identified on 20 chromosomes, including 34 stable QTNs repeatedly detected by multiple methods or in multiple environments and four QTNs with an additive by environment interaction effect. In the decay regions of 34 stable QTNs, three genes related to NFSP in soybean were screened and verified by haplotype analysis, namely, Glyma.13G105400, Glyma.04G063700, and Glyma.04G063100. Multiple regression analysis of 89 QTNs on NFSP was used to establish molecular-assisted selection models for five environments, which could explain 44.82%-55.06% of the phenotypic variation in NFSP. Based on this model, 153 breeding schemes were selected for five environments, which could achieve the breeding goals of NFSP over 43, 43, 25, 31, and 40 under 21AC, 21XY, 23WD, 23QQHE, and 23AC environments, respectively. These results laid the foundation for understanding the genetic mechanism of pod formation in soybean and molecular breeding of high-yielding varieties.